vulnerability-disclosure
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is purely instructional and serves to guide the agent through legal reasoning and documentation processes for vulnerability disclosures.
- [SAFE]: The skill references established German and European legal databases (e.g., gesetze-im-internet.de, dejure.org) for normative verification, which is standard practice for legal research tools.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied data regarding specific software vulnerabilities and case facts. However, it lacks any executable tools or network-active capabilities that could be exploited by malicious inputs, rendering the surface benign.
- Ingestion points: Case details and vulnerability descriptions provided during the intake phase (SKILL.md).
- Boundary markers: None identified.
- Capability inventory: No file system, shell, or network capabilities are utilized.
- Sanitization: None identified.
Audit Metadata